MANUAL FOR INVESTIGATION OF
H Y D R O L O G I C A L PROCESSES IN
MANGROVE E C O S Y S T E M S
B j ö r n K j e r f v e
B a r u c h I n s t i t u t e f o r M a r i n e B i o l o g y and Coastal Research D e p a r t m e n t o f G e o l o g i c a l Sciences
and M a r i n e Science P r o g r a m U n i v e r s i t y o f South CaroHna C o l u m b i a , SC 29208, U . S . A .
A p r i l 1990
Supported by the U N E S C O / U N D P Regional Project "Mangrove Ecosystems in Asia and the Pacific
T A B L E O F C O N T E N T S 2 P R E F A C E 5 1. I N T R O D U C T I O N 2. T H E M A N G R O V E E N V I R O N M E N T 2.1. Distribution of Mangroves ^ 2.2. Climatic Influences g 2.3. Temperature ^ 2.4. Rainfall and Evapotranspiration ^
2.5. Climate Types 2.6. Substrates
2.7. Winds and Weather
3. G E O M O R P H O L O G I C A L S E T T I N G OF M A N G R O V E ECOSYSTEMS j 3 3.1. Deltas 14 3.2. Mangrove Settings 3.3. Estuaries 3.4. Coastal Lagoons 3.5. Coastal Waters 20 4. P H Y S I C A L PROCESSES 20 4 . 1 . Sea Level Variations 22
4.2. Circulation 27 4.3. Material Transport 29 4.4. Classification 31 5. S T U D Y D E S I G N 31 5.1. Philosophy 31 5.2. Sampling Stations along an Estuary
5.3. Sampling Stations across an Estuary
5.4. Number of Sampling Depths ^3
5.5. Sampling Frequency 33 5.6. Samphng Duration 34 5.7. Samphng Procedure 6. G E O G R A P H I C C H A R A C T E R I S T I C S 36 6.1. Drainage Basin 3g 6.2. Mangrove Area 37
6.3. Topography and Bathymetry ^ 9
6.4. Hypsometric Characteristics
40 7. F R E S H W A T E R B U D G E T S
7.1. Importance of Fresh Water Input 7.2. Fresh Water, Material Input Rates 7.3. Chmatic WaterBalance
8. H Y D R O G R A P H I C M E A S U R E M E N T S 8.1. Initial Consideration
8.2. Water Elevation 8.3. Current Velocity
8.4. Temperature, Conductivity, Salinity, Density 8.5. Total Suspended Solids (TSS)
9. D A T A A N A L Y S I S 9.1. General Comments 9.2. Tidal Analysis
9.3. Interpolating Vertical Data Profiles 9.4. Net Discharge Computations 9.5. Net Flux Computations
9.6. Cross-sectional Area Weighting
10. A C A S E S T U D Y : K L O N G N G A O , R A N O N G , T H A I L A N D
(Co-authored by D r . Gullaya Wattayakorn) 10.1. Project Description
10.2. Design and Measurements 10.3. Sea Level Variations
10.4. Hypsometric Characteristics 10.5. Freshwater Balance
10.6. Constituent Concentrations 10.7. Material Flux Computations 10.8. Estimation of Litter Flux 10.9. Final Comments
2.4. Rainfall and Evapotranspiration
The second critical factor is the ratio of rainfall to evapotranspiration. Although mangroves are found in both humid and arid climates, mangrove growth and species diversity is generally greatest m hurnid equatorial areas where rainfall is plentiful and evenly distributed throughout the year Such conditions exist in the humid equatorial regions, e.g the Sunderbans of Bangladesh and India Malaysia, Indonesia Papua New Guinea (Blasco, 1984), and the Pacific coast of Colombia (Snedaker, 1984), where some of the largest single areas of mangroves stands are found.
According to Walter (1977), mangrove ecosystems exist mainly in four of nine main climate types, as determined f r o m over 8,000 climate diagrams collected in the Klimadiagramm-Weltatlas (Walter and Lieth, 1960-1967). O f these four types, the greatest abundance of species is in the equatorial zone o Asia between 10°N and 10°S, where annual precipitation is high and the annual temperature range is small, i h e second most prevalent region is the tropical summer rainfall zone (to 25-30°N and S), where seasona gradients in both rainfall and temperature are well defined. A smaller number of species exist m parts of The subtropical dry zone, where daily temperature ranges are large and annua precipitation is less than 200 m m . Several species are also found on the eastern border of continents in the warm temperate zone, where winters are mild and humidity is high. Walter (1977) suggested that these climatic restrictions are due to the sensitivity of mangroves to frost, but more likely, they are due to warm ocean boundary currents.
Rainfall does not appear to limit the global distribution of mangroves because they grow in arid desert climates as well as wet (Galloway, 1982). Rainfall does, however, have a significant mfluence on the distribution and zonation of species. Rainfall serves as a primary control of salinity in mangrove environments. Several species of mangrove arc facultative, i.e. they appear to survive and even flower and f r u i t , but do not establish permanent groves in freshwater. Avicennia marina and AegialUis angulata appear to be the most salt resistant species, with the upper limit being 90 ppt in interstitial water (Macnae 1968- U N E S C O , 1987a), but lower salinities in tidal waters. Rainfall and accompanying surface and through-flow of freshwater to tidal swamps leaches the soil of excess salts and thereby helps to keep soil salinity within a tolerable range (Oliver, 1982).
Rainfall leaching can also cause substantial nutrient loss f r o m mangroves (Boto, 1982). The degree of nutrient loss from mangrove plants appears to vary with plant type as we I as season. Tuk^V 197U provides a good overview on qualitative and quantitative aspects of macrophyte leachates, and a study on the leaching effects of rain on a mangrove canopy in Southern Florida has been done by Lugo and Snedaker (1973).
Rainfall periodicity is yet another critical factor. I n climates that are humid throughout the year, soils are continuously leached of salts by heavy but evenly distributed rainfall. Salinity levels are general y constant and stable throughout the year. The greatest diversity of species exists '"^hese regions wh.ch a e mainly found in Malaysia, Indonesia, and Papua New Guinea (Macnae, 1968; U N E S C O , 1987a)^ But in arid climates or monsoonal areas with strongly seasonal rainfall distribution, the low rainfall or drought periods lead to high evaporation rates, and consequently, increased sod salinities During the rainy season, this situation reverses, and soil salinity drops considerably. Mangrove spec.es diversity and and monsoonal regions is thus low because of the requirements for tolerance of high «.^^mties and large salinity fluctuations. Dying mangroves are commonly observed around bare areas m the trop.cs tha experience high evaporation (Galloway, 1982) because of increased interstitial sahmties (Macnae, 1966, Spenceley, 1976).
(Blasco, 1984). I n humid climates, rainfall leach'ïs the soil more effectively in the landward zone, whereas on the outer fringe, salinities are elevated by frequent periods of tidal inundation. The less salt-resistant species thus exist in the landward zone, occasionally in association with freshwater marsh vegetation, such as Pandanus. But in arid climates subject to dry seasons or long periods of drought and little river flow, high evaporation rates increase salinities on the landward margin, while salinity concentrations along the outer edge are moderated by tidal inundation. The most salt-resistant species are thus found in the landward zone, often alongside extensive areas of bare sand with hypersaline soil conditions (Blasco, 1984). Bare salt flats exist, f o r example, in northeastern Queensland, Australia (Oliver, 1982) and Tanga in East A f r i c a , where soil salinity can range f r o m 25 ppt at the seaward edge to 41 ppt at the landward zone (Walter, 1973).
Because mangrove zonadon and distribution is greatly influenced by rainfall and evaporation rates, U N E S C O (1979) developed the Map of the World Distribution of A r i d Climates, which classifies climates according to degree of aridity. The common climatological method for expressing aridity is to compute the difference between mean annual rainfall in a region, expressed as P, and the mean annual potential evapotranspiration, PET. When PET is greater than P, i.e., when water loss is greater than water received, a water deficit results. U N E S C O (1979) selected Penman's (1956) procedure for computing potential evapotranspiration, although there are several alternative methods available (Thornthwaite and Mather, 1957; Papadakis, 1965; Ture, 1954). Blasco (1984) provides a comparison of Penman's technique to several other methods of computation.
According to the U N E S C O (1979) classification map, four degrees of aridity were established based on (1) the ratio P/PET; (2) temperature regimes; (3) drought periods; and (4) temperature of the coldest month. I n comparing the U N E S C O (1979) aridity map to that of the worid distribution of mangrove ecosystems, Blasco (1984) concluded that (1) 90% of the worid's mangroves are found in humid regions where P/PET > 0.75; (2) mangroves are occasionally found in sub-humid climates where 0.50 < P/PET < 0.75, such as Kenya, Tanzania, Venezuela, Australia, and Mexico; (3) very few mangroves are found in semi-arid conditions where 0.20 < P/PET < 0.50, mainly only in Ecuador, the Indus delta in Pakistan, Gujarat in India, and parts of Australia; and (4) mangroves are almost non-existent in arid climates where 0.03 < P/PET < 0.20, except around the Red Sea, the Persian Gulf, and Gulf of California. Blasco (1984) does point out several weaknesses in this provisional methodology for classifying mangrove ecosystems, principally among them the lack of systematically collected climatological data and refined techniques for data synthesis. A l l climate classification schemes thus far developed are based on data largely obtained f r o m meteorological stadons and averaged over some twenty years or more. These data suppress yeariy climatic variability and stability measures, which are essendal to mangrove distribution and zonation. More emphasis needs to be placed on the systematic coUecdon of climatological data, particulariy temperature and rainfall measurements f r o m localized areas, to develop an improved scheme for classifying mangrove ecosystems as a means of understanding and managing mangrove wetlands. 2.6. Substrates ' . .
Heavy rains and subsequent high river flow carry alluvial sands and muds to tidal flats, thus forming a substrate f o r mangrove colonizadon and succession. Specimens of some species of mangroves do grow on sand, gravel, or rock shores, but these substrates are generally abrasive and cause considerable damage to most'seedlings (Bird and Rosengren, 1986). Mangroves typically colonize on fine-grained alluvial muds and sands deposited f r o m river and surface runoff onto deltas, inlets, estuaries, lagoons, and other protected regions of the coastal plain (Thom, 1984). Classic examples of mangrove colonization and succession can be seen along the west coast of Malaysia, where extensive mudflats are accreting seaward (Coleman et al., 1970; Diemont and van Wijngaarden, 1975); in the Sunderbans of Bangladesh (Vannucci, 1989); in Borneo (Anderson and Muller, 1973); and along several lagoon shores of Tabasco, Mexico (Thom, 1967).
The accretion of alluvial muds by rainfall can be a slow, long-term process, as would occur m a humid environment with evenly distributed rainfall and continuous deposition of muds. But it can also be caused by short-term (1-2 days) catastrophic events such as flash flooding. I n this case, huge quantities ot alluvium can be deposited over a brief period, often causing initial destruction to mangrove stands before rebuilding would take place.
2.7. Winds and Weather A- ,u Wind affects mangrove establishment in many ways. Local winds increase evaporation rates, aid in tne dispersal of seedlings, and modify the strength of tidal currents and longshore d r i f t , which can mcrease deposition of sands, silts, and clays along coastlines. Heavy sand deposition can bury pneumatophores and existing stands of mangroves. This has been observed on the southwest coast of Madagascar (Blasco, 1984). More often, however, deposition of muds and silts by strong wave acdon and longshore currents provides a substrate f o r more mangrove colonization, as can be observed in Cairns Bay during high spring Udes ( B i r d , 1972).
Stocker (1976) defined four types of physical damage to mangroves caused by wind and waves: (1) windthrow; (2) crown damage; (3) bole damage; and (4) death. The suscepdbility or resistance to each type of wind damage is species specific (Stocker, 1976). Because of the damaging effects of high winds and wave action, mangrove stands are generahy located along low energy coastlines, f o r example, the leeward side of islands or in sheltered areas such as lagoons, estuaries, inlets, river deltas, and coastlines protected by barrier reefs.
Few studies exist on the effects of wind on mangrove physiology, mainly because of the lack of systematic climatological data f r o m localized mangrove communities. Prevailing wmd speeds and directions are frequently available f r o m meteorological stations, but these data are not necessarily representative of conditions within specific mangrove canopies, which have highly localized microch-mates. Emphasis should thus be placed on collection of climatological data directly from within mangrove canopies f o r more accurate studies on mangrove physiology and response to wind and other chmatic factors.
The effects of natural weather disturbances (tropical storms, hurricanes/cyclones, lightning, etc.) on mangroves is another area lacking in available data and references. According to Johns (1986) the majority of case studies dealing with hurricane or cyclone effects on mangrove stands are focused on Central America. But results f r o m these studies are probably not representafive of mangrove stands in general. The ability of certain species to withstand hurricane, typhoon, or cyclone damage is regionally highly variable. For example, Jennings and Coventry (1973) and Chapman (1976) both found Rhizophora to be highly resistant to hurricane damage, but Spenceley (1976) observed that Rhizophora suffered the most damage f r o m a tropical storm in Townsville, Australia (cf. Oliver, 1982).
Stoddart (1962) studied the effects of Hurricane Hattie in October, 1961, on mangrove stands in British Honduras. He reported that defoliation oi Rhizophora, Avicennia, Lagunculana, and Conocarpus occurred over a 20-25 mile zone north and south of the storm track. Those mangroves exposed to flooding and high wave acdon as well as wind gusts suffered more defoliation than those exposed to high winds alone (Stoddart, 1971). A survey four years after the hurricane showed that most of the defoliation resulted i n mangrove death, with no evidence of recolonization (Stoddart, 1965). In most cases, recolonizadon does not occur on dead mangrove stands (Stoddart, 1971).
The influence of lightning strikes on mangrove communities has also received little attention (cf. Johns, 1986), although lightning strikes arc common in the tropics and have caused considerable damage to mangrove stands in Ponape (Sturges, 1865) and potendally long-term commumty restructure to
mangrove stands in Papua New Guinea (White, 1975; Johns, 1981).
Solar radiation influences the growth of mangroves as well. Requirements for light and shade vary among mangrove species, and frequently change during a plant's life cycle (Saenger, 1982) Seasonal fluctuations in mangrove productivity appear to be influenced by seasonal variations in solar radiation (ct. Clough et al 1982), although little research has actually been done to this effect. But solar radiation, wind, temperature, and rainfall act together to influence humidity and evaporation rates in mangrove communities.
Periodicity and duration of data collection are critical factors. Most existing data sets in mangrove environments are based on weekly, or more usually, monthly mean values. The frequency of climatic measurements needs to be increased to at least twice daily sampling, with measurements collected during both day and night. Also, the raw daily measurements should be analyzed as well as the monthly or annual mean data. The daily measurements provide excellent information on stability of mangrove environ-ments, while the mean monthly data can be used to analyze long-term overall climatic effects.
3 . G E O M O R P H O L O G I C A L S E T T I N G O F M A N G R O V E E C O S Y S T E M S
3.1. Deltas . . . . . . .
Mangroves exist under a wide range of climatological settings, but their geomorphological setting is far more restrictive. I t is largely controlled by substrate type, degree of shelter from wmd and wave action, amount of river runoff or surface through-flow, tidal inundation, intersdtial salinity, and soil type.
Deltas are by far the most common geomorphological setting for mangrove colonization. There are several reasons for this. Because mangroves typically colonize from the mid to high tide zone, low-lying, broad deltaic plains subject to extensive tidal inundation provide the largest reaches for mangrove colonization. Although tidal range has often been cited as a major control over mangrove distribution and spread (Macnae, 1966; Galloway, 1982), some of the most extensive mangrove stands in the world are located in deltaic regions that exhibit a low to moderate tidal range of 2 - 4 m , e.g., the upper Gulf of Thailand. Mangroves also exist along coasts that have insigniflcant tides (cf. Lugo and Snedaker, 1974). According to UNESCO (1987a), freshwater discharge is probably a more critical factor in controlling mangrove producdvity than is tidal range.
Although mangrove species vary in their salinity tolerance, salinity is not necessarily a major influence over mangrove zonation. Most mangrove species are facultative halophytes and thus do not require saline condidons for survival. Rather, they prefer brackish to saline waters at least part of the year, maybe because salinity eliminates competition f r o m glycophytes (Thom, 1967). Freshwater discharge or terrestrial runoff, on the other hand, greatly favors mangrove colonization; it supplies nutrients and leaches the soil, thus keeping soil salinity within a tolerable range.
Frequency and range of tidal inundation are critical factors in arid regions with low annual or only seasonal freshwater flow and high rates of evapotranspiration. Under these climatic conditions, bare hypersaline flats are frequently found just landward of the high tide level. These flats rarely support vegetation and are only inundated by an occasional extreme spring tide, which decreases evaporation and temporarily alleviates hypersaline soil conditions.
Deltas are typically prograding because of the continuous supply of alluvial deposits from river discharge. This also favors the colonization of mangroves. Mangroves contribute to land building because they colonize a suitable substrate first and add to progradation once they have been established. Examples of mangrove succession along actively prograding shores can be seen along the coastlines of west Malaysia (Coleman ei al., 1970), the Sunderbans, and lagoon shores of Tabasco, Mexico (Thom, 1967). Fine-grained alluvial silts and sands deposited into deltas are also ideal substrates for mangrove colonizadon. Soil substrate acts as a major geomorphological constraint in mangrove development and zonation (Thom, 1967). Although some mangrove species do colonize on gravel or rocky substrates, their growth is stunted to varying degrees because of the sheer lack of soil, as well as the abrasiveness of the substrates to the roots of mangroves. Exposed beaches are also ill-suited for mangrove growth because marine sands deposited by wind and wave action cover and drown mangrove pneumatophores. Mangrove establishment is optimal in muddy sediments with a high silt content, such as the alluvial muds carried down by river or freshwater surface runoff.
Deltas are typically low-wave energy environments, and thus provide mangroves with protection against harsh winds and waves. Wave action can be highly destructive to mangrove roots that run in the surface layer of the substrate, and can cause erosion where mangroves are growing.
their geomorphological setting varies greatly, depending on the physical process doinmatmg the system. Deltas can be located along low wave energy coastlines with negligible tides and high river discharge, or high wave energy coastlines, or coastlines with extreme tidal ranges. Because each physical process gives rise to different landform types, soils, and biological and ecological characteristics (cf. Coleman 1976), Coleman and Wright (1975) classified deltas according to their dominating physical process: (1) wave dominated, (2) ddally dominated, or (3) river dominated.
3.2. Mangrove Settings „ . . , . • r- i Thom (1982) modified Coleman and Wright's (1975) classification of deltas by describing five general
environmental settings in which mangrove colonization often occurs ( c f . Thom, 1984). However, this classification may not be well suited in all cases and should be used cautiously.
The first setting consists of river dominated deltas along coastlines of low tidal range. Rapid deposition of sands, silts, and clays causes progradadon over a flat, gently sloping condnental shelf. Wave energy along the shoreline is low. Because of high freshwater discharge, the acdve distributaries may not themselves be inhabited by mangroves; however, chenier plains, which form m the margins of these distributaries due to mud deposition f r o m longshore drift, are ideal sites for mangrove colonization. River dominated deltas are also characterized by rapid habitat change and morphologically diverse flora and fauna (Thom, 1982). The Mississippi, the Ganges-Brahmaputra, and the Orinoco deltas are examples ot this f o r m of deltaic plain.
The second setting is ddally dominated. The main distributaries are fed by numerous ddal creeks and are usually funnel-shaped with broad interddal shoals throughout the mouth. Mangroves colonize along the shoals and the shores of the main distributaries and tidal creeks. Examples of this setting are the Klang delta in Malaysia and Ord River delta in Australia (Thom, 1982).
The third setdng is wave dominated with relatively low river discharge. High wave energy results in a more steeply sloped inner continental shelf than in the other settings. The wave dominated setting is characterized by bay barriers or barrier islands that enclose either drowned river valleys or lagoons, and
Lagoon
D e l t a
act to dissipate wave energy. Mangroves are found on the protected leeward side of the barriers and along the shores of the lagoon or drowned river valley. Examples are Laguna de T é r m i n o s in Mexico and the barrier coastline of E l Salvador (Thom, 1982).
The fourth setting is a combination of wave and river dominated processes, with a coastal plain characterized by sand beach ridges and narrow coastal lagoons. Mangroves colonize along abandoned distributaries, near river mouths, and along lagoon shores. A good example of this setting is the Grijalva delta in Mexico and the Purari delta in Papua New Guinea.
The f i f t h setting is a drowned river valley system with low river discharge, low wave action, and low tidal range. Sediment deposition is minimal, creating an open estuarine system. Mangroves colonize the heads of drowned tributary valleys and along the shores of lagoons behind bay barriers near the estuary mouth. A n example of this setting is Broken Bay, Austraha (Thom, 1982).
I n South Florida, Lugo and Snedaker (1974) recognized five basic mangrove community types distinguished by their hydroiogical and tidal characterisdcs. These are: (1) the frmge forest, (2) riverine forest (3) overwash forest, (4) basin forest, and (5) dwarf forest. This classificadon scheme has also been applied to mangroves i n Puerto Rico, Mexico, and Central America. Its limitation is that mangrove community structure in South Florida differs f r o m that in other parts of the worid, for example, Malaysia or Papua New Guinea. This classificadon may not be well-suited f o r all regions.
Davies (1973) viewed coastal environments as part of a continuum of geomorphic types f r o m deltas via estuaries to lagoons (Fig. 1). A t one end of the spectrum exist coastal lagoons, which are located behind wave-built barrier systems and are characterized by sand sediments. Good examples of this type of environment are the lagoons of the Mexico Gulf Coast (Lankford, 1976). A t the opposite end of the spectrum lie deltas, which are river dominated, protrude into a receiving basin, and are characterized by fine-grained sediments derived f r o m terrestrial run-off. Between lagoons and deltas are estuarine lagoons estuaries, and estuarine deltas, represendng a mixture and gradadon of the two extreme coastal
E S T U A R I E S 1. D R O W N E D ^ ^ . y ^ R I V E R V A L L E yS F J O R D 3. C O A S T A L L A G O O N ^ ^ ^
Fig. 2. Schematic diagram of major estuarine types (cf. Kjerfve and Magiil, 1989).
environments. Presumably, a decrease of wave energy coupled with an increase of river sediments shifts a system f r o m the lagoon extreme toward the delta extreme. .
3 3 Estuflrics
' 'Mangrove wetlands are often associated with estuaries and deltas. K j e r f v e (1989) developed a functional definidon for estuaries which includes the main geomorphic types: (1) drowned river valley or coastal plain estuary, (2) f j o r d , and (3) bar-built estuary or coastal lagoon (Fig. 2). Although mangroves do not exist adjacent to f j o r d systems, they are often associated with estuaries and lagoon systems^ K j e r f v e (1989) defined an estuarine system functionally as a pardal coastal indentation with a restricted ocean connection that remains open at least intermittently (Fig. 3). The estuary can be divided into three regions- (1) tidal river zone, a fluvial zone characterized by lack of ocean salinity but subject to tidal rise and fall of water level; (2) mixing zone, or estuary proper, characterized by water mass mixing and existence of strong property gradients reaching f r o m the tidal river zone to the seaward mouth of the system which is best defined by the location of a river-mouth-bar or ebb-tidal-delta; and (3) nearshore zone, a turbid region, usually in the open ocean seaward of the mixing zone but landward of the main tidal front' defined by the extent of the ebb tidal plume.
The boundaries of the three zones are dynamic and change positions continuously, on time scales ranging f r o m geologic to less than a tidal cycle. The landward extent of the ddal river zone can be expected to move downriver with increasing fresh water discharge and change in tidal range f r o m spring to neap The interface between the tidal river and mixing zones will oscillate over a tidal cycle and move seaward with increasing river runoff. The interface between the mixing and nearshore zones will change much more slowly, usually on dme scales longer than the seasonal cycle, but more dramatically over thousands of years. A severe storm could, however, breach a barrier island or reef (see Hayes, 1978) and drasdcally relocate this interface overnight. The seaward boundary of the nearshore zone will change
ESTUARY-COASTAL
S Y S T E M
RIVER (NO TIDE)
NEARSHORE " / Z O N E / C O A S T A L y BOUNDARY
LAYER
O F F S H O R E Z O N E
Fig. 3. Functional regions in a hypothetical estuary (Kjerfve, 1986).
positions depending on the stage of the tide, river discharge, and prevailing oceanographic and meteorologie conditions.
In a given system, all zones may not be present. For example, lagoons in arid or semi-arid coastal regions with a small tidal range may not exhibit a tidal river zone, such as in the case of several hors along the Makran coast of Pakistan. Similarly, an estuarine/lagoon system may not exhibit a mixmg zone it a major river debouching into the estuary happens to be in flood. I n such a case the tidal river could border direcdy on the nearshore zone, and the estuarine mixing processes would then take place within the nearshore zone. This situation occurs in the Amazon River (Gibbs, 1970). Finally, the nearshore zone may not exist in lagoons where the tidal range and river discharge are small, such as m Cancun Bay.
Mangroves are generally most common in the highly protected intertidal regions of deltas, estuaries, and coastal lagoons, corresponding to Kjerfve's (1989) zones 1 and 2 (Fig. 3). To a lesser extent they colonize shorelines associated with carbonate reef; however, in such locations mangrove stands are found in protected areas behind a fringing reef, which dissipates strong wave action.
^''^CorsTaUagoo^cover 13% of coastal environments on a global basis (Barnes, 1980) and represent a type of coastal system different f r o m coastal plain estuaries, although both can conveniently be classified as estuaries. They are pardcularly common in tropical settings and are often associated with fringing mangrove systems, e.g. Laguna Joyuda, Puerto Rico. K j e r f v e (1986a) « f " ' ^ ' ^ ' ' ^ f ^ . ' ^ g ^ f l ' ^ ^ " ^ / ^ ^ f major types according to their degree of water exchange with the coastal ocean (Fig. 4). The rate of oceanic exchange reflects the dominant forcing function(s) and the time scale of hydrologie variabihty.
C O A S T A L LAGOONS
1. C H O K E D ^ ^ - ^ ^ ^
2 . R E S T R I C T E D
3 . L E A K Y
Fig. 4. Main types of coastal lagoons (after Kjerfve, 1986; Kjerfve and Magiil, 1989).
Type 1, examplified by Lake Songhkla in Thailand, are choked lagoons, characterized by a single narrow entrance channel, long residence times, and dominant wind forcing. Type 2, examplified by Laguna de T é r m i n o s in Mexico, are restricted lagoons with two or more entrance channels, a well-defmed tidal circulation, strong wind influence, and usually vertically mixed waters. Type 3, examplified by the Belize barrier reef lagoon, are leaky lagoons, characterized by wide tidal passes, unimpaired water exchange with the ocean, strong tidal currents, and the existence of sharp salinity and turbidity fronts. Mangroves flourish pardcularly in restricted lagoon environments.
3 5» Oo3st3i Wstcrs
" Mangrove distribution along coastlines is influenced by oceanic forcing in the f o r m of tides, waves, sea level fluctuations, ocean temperatures, and structure and bathymetry of the condnental shelf and slope. The portion of the ocean that comes into direct contact with mangrove wetlands is the coastal boundary layer, a nearshore open coast turbid water mass with estuarine characteristics, separated f r o m adjacent condnental shelf waters by a distinct front. A n example of the coastal boundary layer ( C B L ) along the coast of South Carolina, U S A , has been redrawn (Fig. 5) f r o m a Landsat 5 thematic mapper ( T M ) image. The coastal boundary layer usually exhibits high concentrations of sediment, nutrients, and dissolved and particulate organic matter, and is the area f o r spawning of several species of Penaeus shrimp and fish. The dynamic behavior and physical characteristics of coastal waters induce far-field forcing of circulation in adjacent deltas, estuaries, lagoons, and other sheltered coastal features that are colonized by mangroves. For example, estuarine tides are largely forced by ocean ddes at the mouth. Similariy, surface water accelerations in the estuary are forced more by the synoptic wind stress on the coastal ocean surface than by local wind stress in the estuary. This is especially true in the case of narrow, branching, and winding coastal inlets.
Meteorological forcing of coastal and shelf waters by propagating atmospheric pressure systems and wind stress generates condnental shelf waves (Mysak and Hamon, 1969; LeBlond and Mysak, 1978). These are trapped, long waves with periods f r o m 2 to 15 days and wavelengths on the order of 1,000 k m . They travel parallel to the coast on the continental shelf at phase speeds f r o m 0.1 to 1.0 m s-1 (Brooks and
HIGH TURBIDITY
NEARSHORE ZONE
Fig. 5. The coastal bound-ary layer off the coast of South Carolina, USA, from Landsat 5 Thematic Mapper image.
Mocrs, 1977), At the coastline and in estuaries, their associated wave hdgM ^e as great as 0.8 m (Kjerfve et a!., 1982), but heights decrease exponentially away from the coast.
sea level is also often influeneed by nonlinear processes associated with wind-induced surface gravity waves entering the mouth of estuaries and bays.
WhPn hwh ocean waves of long period either break or impinge on an estuarine entrance, wave
(Thompson and Hamon, 1980).
There are several ecological implications of these long-period estuarine events of
in estuaries.
4. P H Y S I C A L PROCESSES
4.1. Sea Level Variations
Coastal mangrove-covered landscapes and landforms exist at their present location as a result of eustatic changes in sea level and global and local land adjustments. During glaciation periods, a considerable fraction of the worid oceans have been frozen into continental glaciers. A t such times, global sea level reached low stands, with coastlines located near the outer edges of the continental shelves. Deltas and estuaries were then probably both small and rare coastal features because of the short time during which sea level remained at any one elevation. Since the onset of the most recent interglacial stage at the beginning of the Holocene period 16,000 years ago, eustatic sea level has risen 130 m at a rapid but variable pace (Nichols and Biggs, 1985) (Fig. 6). I n the process, river valleys and low-lying coastal plain depressions were flooded, river channels were invaded by sea water, and the presently existing coastal plain estuaries and coastal lagoons were formed. Eustatic sea level reached its present elevation approximately 5,000 years ago (Fairbridge, 1980). Since then, small fluctuations in global sea level on the order of a few meters have allowed for the formation of coastal barrier systems due to wave, tide, river, and wind processes. With the flooding of shallow depressions behind the coastal barriers and the condnued accumulation of sediment, coastal lagoons were formed (Lankford, 1976), along with protected salt marsh and mangrove wedands.
Measurements during this century indicate that sea level along tectonically stable coastlines, away f r o m geosynclines, has been rising at a global mean rate of 0.10-0.15 cm/century at the same time that the global temperature has increased 0.4°C ( H o f f m a n et a l . , 1983) as a consequence of a warming chmate and the greenhouse effect.
E U S T A T I C S E A L E V E L
C H A N G E
" 5 0 1 0 0 0 1 0 2 0 3 0 T H O U S A N D S O F Y E A R S B E F O R E P R E S E N TThe relative change in sea level along a particular coast depends on many factors however. Global sea level rise, due to melting of snow and ice, and volume expansion, due to warmmg of the oceans are only two factors. Equally important are (1) changes in the geoid, (2) regional and local tectonic adjustments of the land level, (3) changes in the ocean-atmosphere dynamics (currents, wmds, pressure, runoff, wave and tides), and (4) sedimentation changes (erosion, sedimentation, compaction, land-use dams, and pumping of ground fluids). Relative sea level rise is often accelerated by human activities such as pumping of groundwater, o i l , and gas; construction of dams on rivers; and cutting of ^^^'^'l^^'''^^^^^^ bains. The rapid relative sea level nse of 0.30-0.50 m/century in Bangkok « ^ f ^ l ^ ^ O represent an extreme case of local delta subsidence, in part the result of pumping of groundwate (Milhman, 1988) I n comparison, most of the east coast of North America experiences a more "^^^^f «//^^J^^l^^^^^'J^
0.10-0.40 m/century (Hicks and Crosby, 1974). Palynology, in general, and the f/^tribution of J y ^ pollen, in particular, have been useful for the dating and marking of mangrove shorelines (Thanikaimom, 1987; Vannucci, 1989).
Sea level rise is currently receiving much attention, although equal weight ° " g h t ^o be fo^^^^^^^^^^ regional and local land level change. The present rate of eustatic sea level rise, 0.10-0.15 "^^^"^ury is small compared to the 1.4 m/century rise during the early Holocene. Various scenarios have been modeled to predict sea level rise during the next century as a function of greenhotise ^ ^ ^ / " g ^ D o o m s d a y predictions ( H o f f m a n et al., 1983) go as far as to suggest a possible globa sea level rise of 4.5 m dunng he next century. Although most scientists seem to favor an esdmate of sea level rise of 0.5-1.0 m during the next century, there continues to be much disagreement.
I n comparison, eustatic sea level has remained at an approximate stand-still for Past 5,0(W years (Fig. 6), enabling marine and fluvial processes to fill estuaries with sediment and build deltas and n he proces reshape seaward boundaries of continents (Meade, 1969). The ^^"^^-«'^.^'^«f ^^is sediment mfd^^^^ fs extremely short f r o m a geologic perspective. I n fact, it is so short that estuarine ^eposits laid down m^^^^ than a million years ago are indistinguishable f r o m other shallow-water marine deposits (Schubel and HirschbTrg, 1978). Emery and Uchupi (1972) estimated that i f sea level remained constant and all
A N N U A L S E A L E V E L C H A N G E
E
^ 3 0 1 -LU > UJ OC UJ H <5
2 0 1 0WINTER SPRING SUMMER F A L L
Fig. 7. Typical seasonal mean sea level variation (northern phere |).
sediments with the exception of the Mississippi River load were deposited into the estuaries that presently border the United States of America, these estuaries would be completely filled m 9,500 years. Because all deltas do not build at the same rate, various stages of estuarine and deltaic geological development exist along different coasdines (Schubel and Hirschberg, 1978).
The deltas, estuaries, and coastal lagoons formed during the Holocene eustatic sea level standstill (cf. Fig 6) are the environments that provide the setting for mangroves and salt marshes. Because de tas, esfuaries, and lagoons are more common along coasts with broad and flat, rather than narrow and steep
continental margins (Schubel and Hirschberg, 1978), mangrove wetlands are ^ n n l ' i ; Mangroves are best estabhshed along actively prograding coastlines (Thom, 1984), a direct consequence
of a broad and flat continental shelf margin.
Local relative change in sea level occurs as a result of eustatic sea level change, local subsidence, and tectonic activity. The build-up of mangrove systems, formation of deltas, and shoaling of coastal embayments, estuaries, and lagoons also depend on (1) transport of terrestrial materials froyrodrng continents b^ rivers and (2) net up-estuary transport of sand-s.zed materials f r o m nearsho e waters (Meade, 1969), and are enhanced by deposidon of mud around mangrove roots. As m f i hng and progradadon continue and coastal landforms progress through different stages of geological development mangrove succession occurs with one community preparing the ground for another until a climax community is reached. This is often considered to be a tropical rain forest devoid of mangroves (Thom, 1984).
Not only does mean sea level change on the time scale of glaciation and interglaciadon stages but it also changes significantly over a year (Fig. 7). The annual cycle in mean sea level at ;=°^f ^^^"^^^^^^^^^ stations is largely explained by changes in specific volume (steric changes) as y ^ ^ " ' ^ ^ - ^ ^ ^ ^ f / ^ and winter cooling of shallow waters with depths less than 100 m (Pattullo et al-- 1955). T h . effect i significant even along tropical coasts where the seasonal variations measure 10-30 cm- A tropical locations, however, the temperature effect is less pronounced and seasonal sea level variability s more a function of changes in synoptic wind and atmospheric pressure patterns, changes m the coastal ocean circulation, long-lerm astronomical tides, and seasonal pulses in freshwater runoff, including monsoon effects (Pattullo et al., 1955). For example, extreme seasonal mean sea level changes of 1.65 m occur m the upper Gulf of Bengal due to monsoon runoffs (Pattullo et al., 1955). Semiannual changes m mean sea level (cf Fig 7) also occur as a result of runoff events, as in the case of the Miss.ss.ppi River, U S A , which
produces a springtime high mean sea level. A t other stations the semiannual « « f ^"^^^^^^^ is associated with variability in the large scale ocean basin circulation (Pattullo et al., 955). I t i clear tha
seasonal sea level variability can have a dramatic effect on the flooding of coastal wedands I n the case of North Inlet, SC, tidal inundation of the entire salt marsh is more likely to occur in September than in February, bearing associated ecological impacts ( K j e r f v e and McKellar, 1980).
"^'^ W^ateT'drcuLion in coastal waters and wetlands controls turn-over times and plays an important role in the transport of nutrients, sediment, salt, and other dissolved and suspended materials. Thus, water quality is directly related to water circulation. Also, larval recruitment of many species into estuaries and mangrove or salt marsh wetlands depends on local circulation patterns. I n calculating mass budgets f o r coasfal ecosystems, it is essential to be able to esdmate the import and export of water-borne matenals. This requires a good understanding of the circulation and local mixing processes. But care must be exercised in interpredng estuarine processes. JUst because the circulation causes water transport does no necessarily mean that dissolved and particulate constituents will be transported in the same direction as the water flow, because material transport also depends on local mixing.
I n designing studies of mangrove ecosystem processes, it is essential to have a good understanding of the dominant hydrodynamic time scales. For example, the hydrodynamic residence time a" estuaiy i the water volume of the system divided by the rate of change of volume due to ^^^hwater discharge ti^^^^^ exchange, or water circulation. Comparison of hydrodynamic turnover time to a bjogeochemical m^^^^^^ time indicates whether or not the biogeochemical processes are dependent on ^^e hydro ogica characteristics of the system. For example, if a coastal water body has a ^ r o d y n a m i c turnover U^^^^^^^^ several days, but a phytoplankton populadon has a turnover time on the order of hours, the two processes are not l i n k ; d and can be treated separately. Similarly, when considering the f fimentation rate m he same water body with a time scale of centuries, the coupling between the '^'^^^'''^^'^'^J'^^^^^ can be treated separately. However, when the hydrodynamic and biogeochemica turnover times are similar, the two processes are coupled and it becomes necessary to study the hydrology m analyzing the biogeochemical process.
Circuladon refers to residual or time-averaged water movement. Because l^'^'^^^^^^^^J^}^^^^ continuum of time scales, it is critical to choose an averaging time '^^'^'P'^'^^^'^^^^^^
with the greatest variability in calculadng residual currents. Because coastal and "^/"^^^^^^^^ largely tidal, one or more complete semidiurnal tidal cycles is the ^PP^^P'^^'^.^"^.^^^^^^
can never be determined f r o m a single instantaneous measurement; rather, it is a quantity that must be calculated based on systemadc measurements over an extended time period ( K j e r f v e , 19/V).
The time-averaged currents that constitute the circulation vary depending on locadon within an estuanneTr coastafsystem and the depth at which the measurements are made, t is common practice to refer to these time-average currents as net currents, tidal residuals, or non-tidal flow.
Three forcing functions drive the circuladon in estuarine and coastal waters: (1) freshwater discharge (2) L a i currents' and (3) wind stress. Variabihty in oceanographic conditions - / ^ f
also modify the circulation in many estuaries. This process is referred to as far-field ^^^^^'^g (^^^^^
Wind waves can modify the circuladon in wide or exposed coastal systems as wel I " ^ ^ y f ^^^^ however, the effect of wind waves is of minor importance because mangroves only grow well where wind waves are absent.
Geometry and bathymetry of coastal systems, bottom fricdon, and ^«^ation of the earth (Coriolis acceleration) do not drive currents direcdy, but can sign ficantly alter currems in enclosed water bod es^ These effecis become increasingly important in modifying /^e circulation as the m^^^^^^^^^^
instantaneous currents becomes greater. Only when currents already exist do these fac^^^^^^ t n afso aUer Man-made alterations such as dredging, channelizadon, damming, and channel diversion, can also alter circulation patterns in coastal water bodies.
Each one of the three forces drives a pardcular kind of water circuladon: (1) freshwater discharge i n d u c e s \ r i t a d o n a / chc^ (2) tidal'currents drive a ^^^^^^^f^^'^^^^^^^^
stress causes wind-driven circulation. Although an estuary or coastal water body is dominated by one or hese c S ^ r o n typ^^^^ or all three types could be operadng simultaneously in the same estuary, Resulting in a complex flow structure that is often difficult to interpret f r o m measurements.
Gravitadonal circulation is caused by density differences between f^^^^water runoff an^^
ocean water Freshwater runoff, being less dense, has a tendency to remain mostly m the surface layer Though t^^^^^^^^^^^ tend t'o mix the water column, ^ s turbulent mi^^^^^^^^^^^^
vertical exchange with high salinity water f r o m the bottom layer mixing into the surface layer, ana low a ï r y wa e '^^^^^^ surface mixing into the bottom layer, As a result longitudinal and v e " - ty and density gradients are formed in coastal water bodies. These density gradients correspond to
time-averaged pressure gradients, wliich drive tlie gravitational circulation.
The pressure surfaces along the main axis of an estuary tilt seaward in the surface layer, causing a net outflow, and up-estuary in the bottom layer, causing a net landward flow. A t mid-depth, the pressure surfaces become horizontal or parallel to an equipotential surface, upon which the gravitational potential is constant. Where the pressure surfaces become horizontal, the net flow vanishes. This level of no motion usually slopes slighdy across an estuarine channel as a result of earth rotadon (Pritchard, 1952 and 1956) or channel curvature (Stewart, 1957).
The main circulation induced by the mixing of fresh and ocean waters in estuarine systems is the classical estuarine circulation, a form of gravitational circulation (Pritchard, 1956; Pritchard and Kent, 1956; Dyer, 1973). Analydcal similarity solutions of the dynamic equations have been obtained for the two-dimensional case, ignoring changes occurring across the estuary (Rattray and Hansen, 1962; Hansen and Rattray, 1965). These solutions provide a f i r m theoretical basis f o r understanding the physical dynamics of estuarine systems in which an upper and lower layer are present.
Gravitational circulation also occurs in coastal waters that are well mixed vertically by winds, waves, or bottom-generated turbulence. I n such systems, horizontal differences in salinity, and therefore density, still cause a seaward pressure gradient and associated net seaward flow.
The amount of water transported as part of gravitational circulation is always greater than the amount of freshwater discharge. For example, i f river discharge into an estuarine system measures R units (in m^ s->), the surface outflow at the mouth of the same estuary can be 25 R units (Schubel and Pritchard,
ESTUARINE GRAVITATIONAL
CIRCULATION
L A N D O C E A N
Fig. 8. Schematic estuarine gravitational circulation (after Kjerfve, 1989).
F L O O D E B B F R E S H S A L T Y
1972) , implying that 24 R units enter the estuary in the bottom layer. This concept is diagrammed m Fig. 8 (cf Dyer, 1973; K j e r f v e , 1989). Obviously, the flushing time is significantly shorter for a system with a weil developed gravitational circulation. A litde bit of freshwater discharge into an estuarine system goes a long way in driving gravitational circulation.
The resulting salinity distribution (cf. Fig. 8) is primarily responsible for the de"«i;y f ^dients^^^I^^^ embayments and estuaries, salinity differences have a much greater influence on '^'^^"l^^ion^^^^^^^^
temperature or pressure changes (see K j e r f v e , 1979). Temperature differences
'^J'^'J^^TL^Z oi
gravitadonal circulation in lagoons that receive little or " « / y ^ ^ h w a t e r runoff Intense so ar he^^^^^^^^ shallow lagoons somedmes causes excessive evaporation and, because of ^ - ^ f Pf^^^f^^^^J^^^^^^^^^^ super-elevated salinities in the interior of the lagoon (Collier and Hedgepeth 1950). I " ™ f ^ / ^ ^ ^ ^ ^ ^ ^ ^ ^ world, e.g. northern Australia, there are large fluctuadons in the seasonal discharge. ^ i and R i ^ ^ (1986) and Wolanski (1986, 1988) observed that during the dry season, there exists a^f^'^y^^''^'^^
zone in many tropical Australian estuaries. The salinity maximum is often a ew PP "^^^^^^^^^^ coastal salinilty and is located halfway between the head and the mouth of estuary becam combined evaporadon f r o m the water surface and evapotranspiration from the adjacent ^^^^^^"^^^^^^^^^exceeded the runoff. Apparently this situation occurs commonly in many tropical „ 7 " ^ ^ ^ ^ ^ systems and gives rise to a diffusive, ocean-directed transport of salt (Wolanski and Gardiner 1981)^^^^^^^^^
upstream salinity maxima could also give rise to an inverse gravitational circulation (Wolanski, 1988) and would certainly have imphcations for mangrove development.
Whereas ocean salinity is 35 ppt, lagoons may commonly exhibit salinities in
'^I'^^'^J^^^^^^^
1973) or higher. This is especially true during the summer in and or «^"^^:^"V w T L a n c inflow Hedgepeth (1950) suggested that this process sets up an inverse estuarine circulation, with oceanic inflow i " a fuTface l a y e r i n d ' o u t f l o w of dense saline lagoon water in a bottom layer. I t J ^ ^ ^ ^ ^ ^ ^ ^ whether this is a reasonable hypothesis. Lagoons are often very shallow and °P^^." f ' ^ Z ^ "become vertically mixed, particulady by wind. Complete vertical mixing ^"hibits development o two-layered flow, inverse or otherwise. Rather, salinity or density dif erences are manife ted a^ differences betwe;n ebbing and flooding flow and across the channel rather ^^an whh def^th. T ^ ^
to a combination of tides and winds as being more important in driving circulation in lagoons. Residual estuarine circulation is also caused by instantaneous tidal currents.
modified by the bathymetry and geometry of the embayment or estuary, ^'^hough the tidal flow is ar^^ oscillatory, it produces a residual tidal circuladon through non-linear
result of bottom friction, shoaling, and changes in width. This usually manifests ^ ^ ^ f ' " ^ j ^^^^^^^^^ in the strength of maximum ebb and flood currents and in the duration of ebbing and flooding tides. When curremfa^e ave^^^^^^ over one or more tidal cycles, the result becomes a net non-zero current induced by the oscillatory tidal flow, and is often stronger than the gravitational circulation.
Non-linear effects in tidal currents occur because of variable cross-sectional width, differences in water deofh e S e L e of tiSal fl^ and channel curvature, all of which can create large spatial velochy t l n ~ i m l r v e r a g e d (residual) tidal currents are often systemadcally e^
of an estuarine cross-section and flood-directed on the other ( K j e r f v e , 1978, K j e r f v e and yroehl, i V / V j (Fig 9) TO Ts frequently referred to as lateral circuladon, and is the result of boundary interaction and chLnnel curvature ?ather than Coriolis effect. Oppositely directed net currents m a cross section do no i m X a nelToss or gain of water in the long term. Still, residual tidal circulation is in many systems responsible f o r systematic export or import of water-borne constituents.
Tidal circulation is pardcularly pronounced in estuaries with shallow water depjh and brge tidal range^ For example, K j e r f v e and Proehl (1979) measured residual tidal flow of 0.5 m s m North Inlet, South
TIDAL RESIDUAL
CIRCULATION
Fig. 9. Tidally driven cir-culation in estuaries (cf. Kjerfve and Proehl, 1979).
Carolina, U S A , and Tee (1976) modeled residual tidal circulation in the Bay of Fundy up to 1 m s"'. Tidal and gravitational circulations co-exist in many systems, although the interaction between these circulation types has received only scant attention. I n shallow systems with a tidal range of 2 m or greater and a moderate to high river f l o w , neither tidal nor gravitational circulations can be ignored.
Wind-driven circulation is particularly important in coastal lagoons. Large expanses of open water, shallow water depth, small tidal range, and low freshwater inflow are conditions that favor dominance of wind-driven currents and water level variations. These currents have not been well-studied because they are highly variable and are often masked by gravitational and tidal flows. Because the wind is variable over a range o f periods f r o m minutes to weeks, it is seldom practical to calculate the circulation associated whh winds. Rather, the wind-driven, instantaneous currents are of prime interest as mixing agents and in causing material dispersion.
A n important energy input to coastal and inshore waters comes from meteorological frontal passages. These can recur every three to twenty days and are most energetic in higher latitudes, but can sdll be important within the distribution range of mangroves. The duradon over which estuarine currents must be averaged to yield the wind-driven circuladon is therefore very long, a multiple of the frontal passage cycle (Weisberg, 1976a; Weisberg and Sturges, 1976). This usually makes it impractical to measure the wind-driven circulation and difficult to assess the exact influence of winds on the circulation. However, variations in mean water levels are cleariy related to meteorological forcing. If the semidiurnal and diurnal water level changes are filtered out, the mean sea level is still quite variable as a result of meteorological forcing (Fig. 10). Although this forcing can be due to local wind stress, it is probably more often a result of far-field forcing of shelf waters by storm systems.
Several empirical studies point to a major response of the time-averaged estuarine flow to wind-forcing. Weisberg (1976b) found that approximately 48% of f^e c ^ ^ ^ ^ ^ ^
Narraganseft Bay was related to meteorological variability on time scales f ^^^^^^^^^^^ Cannon (1978) and Holbrook et al. (1980) showed that even deep currents unlets and
iP^^^^^l
well-correlated with meteorological events. Water surface slopes as large as 4 X l ^ / ^ ^ m ^ s were attributed to 10 m s^ wind forcing in a Louisiana bar-bu.lt estuary ( k j e r f v e 1973) an^m^^^^ found significant water exchange between Corpus Christi Bay, Texas «"'^, h ^ ^ ^ fmeteorological forcing. These wind tides in Texas lagoons have been wel documented (Collier and Hedgepeth, 1950; Copeland et al., 1968). During northeriy winter storms water piles up " the
of the lagoons. This sets up a seiche movement, with water oscillating back and forth "^^^ the length of the lagoon, at a period that depends on the length and depth of the system, ^ ^ t " « " ^ l ^ y j J^^^^^^^^^^^^^ ecological significance of these wind ddes is dramatic-a water level change of only ^^^^^^f^^^^^^^^
expose hundreds of square kilometers of mudflats or inundate an equally large area of coastal salt marshes and grasslands.
Theoretical solutions of the hydrodynamic equations indicate that a steady, along-estuary v^^^^^^^^^ wili significantly augment the gravitational circulation (Rattray and "/"^^"'Z^^^'» ^^^^^^ 1965) The effect of a down-estuary wind is to increase the net surface outflow but at the ^a'"^/^';"^
ncrease the net inflow at depth. On the other hand, in the case of a steady wind blowing into the estuary
Ihe
resuUhig n^^^current
may be flood-directed, a mid-layer would experience net ebb flow, and the net bottom-flow would be much reduced and flood-directed.The circulation in many coastal plain estuaries exhibits modes which ^ " ^ ^ ^ ^ ^ ^ ^ ^ . f ^ ^ ^
These modes were initially not reported in the literature because they were thought of as being just ïnusuaTevïnt:- ( P ; S ^ ^ (1976), however, identified ^ i ; ; ^ different cir^^^^^^^^^^
based on field data f r o m a coastal plain estuary: (1) classical circulation, with surface outflow and bottom
nflfw;
2 v^^^^^^^^^ with'surface inflow and bottom outflow; (3) thre-layered c^^^^^^surface and bottom inflow and outflow at mid-depth; (4) reverse three-layered circu at on v^^^^^ surface and bottom outflow and inflow at mid-depth; (5) discharge circulation, ^^th outflow at all dep^^^^^^ storage circulation, with inflow at all depths. The classical estuarine
''''^^'''^Y^'J^^^^^
mode in the estuary. I t occurred 43% of the time and lasted, on the average, five t dal cycles. The othe modes occurred a smaller percentage of the time and usually lasted for one to four tidal cycles. The mo T m a r i r conclusion of t'his s t u d y l that 57% of the time, ^ ^ - f - ^ - " " ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ p e common circulation mode. Estuarine circuladon is seldom in steady state, ^^^t rather exh bits c^^^^^^^^ temnoral and soatial variabilities. Several independent mechanisms are responsible for altering es uarme S a t l m o d e ï T^^^^^^^ changing wind stress, variations in river discharge varymg tidal range due to fortnighdy spring-neap tidal cycles, and far-field forcing f r o m the coastal ocean.mixing f r o m circulation alone. The advective and diffusive transport components must both be considered to esdmate the transport in coastal water bodies.
Water molecules f r o m ocean and runoff sources cannot easily be differentiated. Thus, it is usually impractical to determine mixing by use of isotopic ratios of stable and radioactive hydrogen or oxygen atoms within water molecules. Rather, the distribution in time and space of dissolved materials in coastal waters is used to determine turbulent diffusion coefficients, which, multiplied by the gradient of the mean concentration, is a measure of the diffusive transport component. The dissolved substance is usually salinity, but could be any naturally occurring conservative and dissolved constituent or a dye or radioisotope introduced f o r the purpose of an investigation. Salinity distribution is most commonly used because (1) salinity is conservative, i.e., salt concentration cannot be altered by biogeochemical processes but only by conservative mixing processes (advection, diffusion, rainfall, evaporation, and freezing); (2) salinity in coastal waters is derived f r o m the ocean source with a salinity of 35 ppt, whereas most land runoff has a salinity less than 0.6 ppt and a different ionic composition; and (3) salinity is measured quickly and easily in the field (cf. K j e r f v e , 1979) and does not have to be determined with great precision on account of large temporal and spatial salt gradients within most coastal water bodies.
The overall transport of water-borne materials can be partitioned in a number of different ways and attributed to (1) river discharge, (2) ddal residual circulation, (3) gravitational circuladon, (4) wind-driven circulation, (5) storage effects, (5) cross-correlation between velocity and concentration (tidal sloshing), (6) vertical and lateral velocity and salinity shear effects, (7) tidal trapping or chopping, and (8) short-term turbulent diffusion. A l l of these processes cannot necessarily be separated f r o m each other. Mathematical decomposition of the transport into these different components and the use of field measurements to determine the relative contributions of the processes to the transport have been carried out by Bowden (1963), Pritchard (1969), Fischer (1972, 1976), Dyer (1974), and Murray and Siripong (1978), K j e r f v e (1986b) and many others.
The process of tidal sloshing refers to the net transport of particles by oscillatory tidal currents over a tidal cycle. I t is usually a major component of the overall transport in shallow tidal systems. When time-averaging the product of estuarine velocity and particle concentration, the integral seldom vanishes because of phase differences between current and concentration time-representations ( K j e r f v e , 1975; K j e r f v e , 1986d). As a result, material transport due to this mechanism is usually transported into estuaries over a ddal cycle.
The shear effect is net transport due to systematic co-variations of velocity and particle concentration over (1) depth (vertical shear; Bowden, 1963); (2) width (lateral shear; Fischer, 1979); or (3) cross- secdon (cross-sectional shear; Hansen 1965). I t is a major inward transport component in coastal systems with strong verdcal gradients. Vertical shear and lateral shear are probably equally important (Dyer, 1974; Fischer, 1976; Murray et al., 1978; and Rattray and Dworski, 1980).
Transport due to tidal trapping occurs when water is temporarily caught in shoreline indentations and branching channels, because the tidal current oscillates past these shoreline features. The trapped water parcels and particles are released into the flow and replaced by new water and particles. The net effect is longitudinal material transport (Okubo, 1973; Fischer et al., 1979).
Mixing intensity and material transport in coastal systems is seldom in steady state over several tidal cycles. The same estuary may at times mix quickly and completely because of a frontal passage or strong winds and at dmes slowly and incompletely because of lack of winds. Similariy, variations in tidal range have been shown to have a profound effect on vertical mixing in the Chesapeake Bay (Haas, 1977). During neap tides with a small tidal range, ddal energy is limited, and the water column becomes
vertically stratified. This stable stratification inhibits vertical mixing. During spring tides, the tidal range and maximum currents are increased, and a sufficient level of tidal energy is available to break down the verdcal density stratification. The result is enhanced vertical mixing, allowing nutrients and food particles in the bottom layer to mix up into the photic surface zone and enhance producdon. Many estuaries behave similarly and alternate between being stratified (limited vertical mixing) and well-mixed (complete verdcal exchange) during the fortnightly spring-neap _tidal cycle (Haas, 1977).
Many water-borne substances, including suspended sediment, settle out in estuaries and other coastal systems f o r a variety of reasons. Coastal systems are often sinks for particular constituents and the estuarine and coastal zone act as a filter. This is largely the case with fine-grained sediments f r o m river sources as well as sand-sized sediments f r o m the coastal ocean (Meade, 1969). Thus, one important question to ask is whether a dissolved or suspended constituent mixes conservatively within the estuary. I t would do so i f the material concentradon changed in proportion to the change in salinity. Salimty is a conservadve consdtuent and i f a material concentradon plotted linearly agamst sahmty, it too would be conservadve. Such a plot of salinity against a material concentradon is referred to as a mixmg diagram. Systemadc deviation of a measured estuarine concentration f r o m a straight line in a mixing diagram is interpreted to imply non-conservadve behavior of the constituent, and this usually implies that the system is a sink f o r a pardcular constituent, or in some cases, a source.
Extreme care must be exercised in using mixing diagrams. The transformation f r o m distance to salinity assumes that (1) measurements are averaged over one or more complete tidal cycles and (2) that concentradons i n the ocean and river ends of the system do not fluctuate i n dme f r o m one tidal cycle to the next (Officer, 1979; Officer and Lynch, 1981; Loder and Reichard, 1981). These assumptions are seldom met in the strict sense. Thus, deviations f r o m a straight line in a mixing diagram may not necessarily mean non-conservadve behavior, which in turn could lead to incorrect transport estimates.
4.4. Classification , . '• Hansen and Rattray (1966) produced a useful dynamical classification scheme of estuarine systems based on two non-dimensional parameters. They classified estuaries or coastal systems m a diagram f o r m (Fie 11) by plotting stratificadon vs. gravitadonal circulation. The stradfication parameter is the ratio between the net bottom to surface salinity difference to the net depth-averaged salinity. The gravitational circulation parameter is the rado between the net surface flow to the freshwater flow. The net surface flow assumes steady state and is taken as a representative value across an estuarine section to smooth out lateral effects. The freshwater flow is simply the freshwater discharge divided by the cross-sectional area (Hansen and Rattray, 1966; K j e r f v e , 1979; K j e r f v e , 1989). The classification diagram can be used to display how particular systems change seasonally. Hansen and Rattray (1966) found that most estuanes could be grouped into four regions on their diagram.
Class 1 estuaries are either coastal lagoons or bar-built systems and are subdivided into two subgroups Class l a estuaries are well mixed vertically and include bar-built systems such as North Inle , South Carolina, ( N I ) and Mississippi Sound. Class l b estuaries exhibit vertical stradfication and include the Vellar estuary, India ( V E ) . Both subclasses lack gravitational circulation. Net upstream salt transport takes place due to turbulent diffusion processes alone, which in this case includes residual tidal flow and wind-driven currents.
Class 2 estuaries include partially mixed coastal plain estuaries and are divided into well mixed (2a) and weakly stradfied (2b) sub-classes. Class 2 estuaries are characterized by a reasonaWy well- developed gravitational circulation and longitudinal transport by both advection and turbulent diffusion transports Some examples are the South Santee River, South Carolina (SS), James River, Virginia (JR), and the Narrows of Mersey, U K , ( N M ) .
E S T U A R I N E
C L A S S I F I C A T I O N
NON-TIDAL S E A L E V E L
CHANGE
W I N T E R S P R I N G S U M M E R F A L L
Fig. 10. Nontidal variations in sea level in a salt marsh system (after Kjerfve, 1989).
10 102 103 104
CIRCULATION Ug/Uf
Fig. 11. Estuarine classification based on circula-tion and stratificacircula-tion (Hansen and Rattray, 1966).
Class 3 estuaries are characterized by strong gravitational circulation and medium to strong stratification. This class includes most fjords, e.g.. Silver Bay, Alaska, (SB) and H i m m e r f j r d , Sweden ( H F ) , and a number of estuarine straits, including the Strait of Juan de Fuca, Washington (JF).
Class 4 estuaries are strongly stratified systems without gravitational circulation. Vertical mixing is caused by entrainment due to breaking internal waves. The lower Mississippi River ( M R ) belongs in this category.
5 . S T U D Y D E S I G N
' • ' v f A c c o m p l i s h a meaningful estuarine hydrographic study in mangrove
^-—^Ssi^^^
carefullv olanned sampling strategy. A n optimum study design depends, however, on the hypotheses to beested and the ov^^^^^^^ thus no simple cookbook recipe for such. This manual ts S n e d l the assumptioi that the study objectives are to (1) describe the hydroiogical c h — a c o L a l system consisting of mangrove wedands, adjacent estuarme water ^^^^^^^^^ coastal marine waters separated from the shelf waters by a sharp water mass front, (2) ""^te^ m^^ baLnces of a mangrove-estuarine system from direct and indirect measurements of wa er-borne materia fluxes t h ^ ^ y T m , and (3) analyze fleld and map measurements to allow a synt esis of the domman hydroiogical processes in support of hypothesis testing and subsequent modehng Thus a ^m^^^^^^ component of the study consists of making hydrographic measurements m the water body bordering the m a ^ g ^ e system. This section is concerned with designing a strategy for carrying out such hydrographic measurements in an estuary or enclosed water body.
Estuaries and other semi-enclosed coastal systems typically experience
^^'f^^^'^f^^^^
variations in material concentradons, water elevation, and flow velocity as a result of t^^^^^^^ influence^ freshwater discharge, meteorological forcing, and basm geometry and geomorphology. Both honzontai and verdcal spadal variations are the norm. For example, the classical estuarine circulation (cf. DyerTo facilitate analydcal solutions to governing equations, estuaries have f ^ ^ " " ^^^^7/^^^^^^^^^^ in a quasi-steady state fashion, implying that temporal 'changes primarily occur on tid^^^^^^^^
scales not exceeding a day. Long-term trends are assumed to be non-existent. This usually doesn t agree wkh melrements ho and coastal water bodies are seldom in quasi-steady state. Even when
due to changing meteorological conditions and river flow variations (Elliott, 1976).
Semi enclosed coastal systems are therefore difficult to study analydcally, which is why it is now
importance and constraint on the field strategy in terms of selecting (1) •«^.^^lon of saniphng statio^^^^^
the system, and the available personnel effort.
' ^ ' ^ ^ ^ l ^ ^ i o . of estuarine processes, d i - c e b e t w e .
main axis of an estuary should be great enough that differences in mean ^'^^^^'''^^^^^^^
